There is a critical need for research to define the social communication phenotype in very young children with ASD. The major aim of this longitudinal study is to quantify social communication skills of children with ASD during the second year of life. Three groups of children will participate in this study, one group with ASD (n=77), one group with developmental delays in which ASD is ruled out (DD; n=77) matched on age, nonverbal cognitive level, race and gender and one group with typical development (TD; n=77) matched on age, race, and gender. All of the children will be recruited from the ongoing longitudinal study of the FIRST WORDS[unreadable] Project. Children will be drawn from a general population sample of over 11,000 children screened with a parent-report checklist. This prospective, longitudinal study will collect videotape samples between 12 and 24 months of age using standardized procedures to gather precise measures of social communication skills. Half of the children will enter the study between 12 and 17 months of age and half will enter between 18 and 23 months. Outcome measures of language, adaptive behavior, and autism symptoms will be obtained at 3 years of age. The following research objectives will be addressed: 1) to compare quantitative measures of social communication of children with ASD in the second year of life to those of children in the DD and TD group; 2) to examine patterns of change in social communication over the second year of life in the subset of children who entered the study between 12 and 17 months of age; and to examine the predictive relations between social communication in the second year and outcome measures at 3 years of age. The expected outcomes of this study will have the following important implications: 1) within- and between-group differences in social communication will contribute knowledge to inform biological research and help to make connections between neuroimaging and genetic findings and behavioral measures; 2) well-defined descriptions of the natural history of social communication and language will inform intervention research by predicting which aspects of behavior are most likely to change in response to intervention and delineating meaningful outcomeswhich can increase statistical power to measure response to treatment; and 3) between-group differences in social communication will indicate more precise early indicators of ASD so that effective intervention programs can be provided earlier.